THE SURFACE TENSION OF LIQUID SULFUR DIOXIDE

surface tension of the liquid to be 33.5 dynes per cm. ... 4 Quinn, This Journal, 49, 2704 (1927). ... The jackets were filled about one-fourth full a...
5 downloads 0 Views 348KB Size
410

Vol. 51

VeRNON M. STOWE

Acknowledgment for generous financial assistance in this research is made to the Carnegie Institution of Washington and to an anonymous benefactor of this Laboratory.

summary No difference was found in the atomic weights of specimens of copper from the Calumet and Hecla mines in the Lake Superior region in Michigan and from Chuquicamata, Chile. The ratio of the atomic weight of copper to that of silver was determined by analysis of pure, recrystallized cupric chloride. The copper was determined electrolytically and the chlorine by nephelometric titration with pure silver. On the basis of Ag = 107.880 the atomic weight of copper was found to be 63.557. CAMBRIDGE, MASSACHUSETTS [CONTRIBUTION FROM THE SCHOOL OF CHEhfISTRY, UNIVERSITY OF

MINNESOTA ]

THE SURFACE TENSION OF LIQUID SULFUR DIOXIDE BY VERNONM. STOWE RECEIVED OCTOBER 1, 1928

PUBLISHED FEBRUARY 5, 1929

The surface tension of liquid sulfur dioxide has been determined once only and that a t a single temperature. At -25’ Grunmach’ found the surface tension of the liquid to be 33.5 dynes per cm. The author has been interested for several years in the work of Dr. Bond2of the University of Iowa on solubility. Kow Hildebrand3 recounts the relation between surface tension and solubility. Therefore, it seemed very desirable to determine the surface tension of liquid sulfur dioxide a t several temperatures in order that comparisons might be made with other liquids. The problem of the solubility relations of liquid carbon dioxide has already been attacked in this same manner by Q ~ i n n . ~

Experimental The capillary tube method was used. A large number of capillary tubes were drawn from a variety of soft glass tubes and of these six were chosen for uniformity of bore, determined by means of a micrometer. They were washed with a mixture of potassium dichromate and sulfuric acid for over twenty-four hours. They were thoroughly rinsed with tap water, then with distilled water and finally with conductivity water. They were dried by drawing air through them briefly and were placed in a desiccator over phosphorus pentoxide for more than a week. From the time they were removed from the cleaning solution they were handled with carefully cleaned hands and only a t one end. This end was broken off immediately to prevent impurities from 1

a

Grunmach, Ann. Physik, [4]4,367 (1901). Bond and Beach, THISJOURNAL, 48,348-356 (1926); various University of Iowa

theses. 3

Hildebrand, “Solubility,” The Chemical Catalog Co., Inc., New York, 1924. 49,2704 (1927).

* Quinn, THISJOURNAL,

Feb., 1929

THE SURFACE TENSION OF LIQUID SULFUR DIOXIDE

411

spreading down onto the cleaned portion. The capillaries were loaded into tubes of heavy pyrex glass 1.35 cm. in diameter, sealed a t one end. These tubes had been thoroughly cleaned and dried in the same manner as the capillaries. The capillaries were held in a position along the axis of the "jacket" tube by means of two platinum wires bent into a bracket shape, with a loop in the middle. The upper part of the jacket tube was constricted t o make subsequent sealing easy. Liquid sulfur dioxide was prepared by the method mentioned by Bond and Beach2 except that no mercury seal was used. The apparatus was well swept out by preparing several portions of the liquid. It boiled a t -9.8" a t 735 mm. pressure, leaving no residue. The jackets were filled about one-fourth full and sealed off a t a length of about 10 cm. A thermostat was constructed of a quart size unsilvered wide-mouthed Dewar flask. This was cooled to -90 and -50" by pouring liquid air into a long test-tube and inserting this tube into the thermostat liquid. Ethyl alcohol was used a t the low temperatures. The alcohol would congeal about the tube and furnish a fairly even temperature for some time as it thawed. Ice and calcium chloride were used for the temperature -20°, and salt and ice for -10 and -5". Ice water and much ice were used for the 0" point. Water a t 4' was used to cool tap water to achieve temperatures between 0 and 25 '. Temperatures of 30 t o 50' were attained by inserting a long narrow lamp commercially used for illuminating showcases. This was turned off and on by means of a home made thermoregulator containing toluene and a mercury contact column. The temperature could easily be controlled within one-tenth of one degree. The light was turned on and off several times per second even when all vibration was eliminated. The bath was stirred efficiently a t all temperatures. The thermometers were checked against one or two standard thermometers recently calibrated by the U. s. Bureau of Standards. At sub-zero temperatures it was necessary to check against freezing and sublimation points. Carbon dioxide, mercury, chloroform, chlorobenzene and carbon tetrachloride were used. Tubes were kept a t the required temperature for thirty minutes to one hour before readings were taken. When working at temperatures somewhat removed from room temperatures, they were placed in a crude cooling bath or warming bath in order that they might come quickly t o equilibrium when used. Readings of the height of the column in the capillary were made by means of a good heavy cathetometer. At least five readings checking within iO.01 cm. were made. These readings were made with the tube suspended in the thermostat in a t least two different positions and a t different levels to eliminate errors due t o distortion of light by the Dewar flask and the jacket tube. Discrepancies due to this cause were rare. The averages of these readings are recorded in Table 11.

After reading the height of the liquid in the capillaries a t the various temperatures the tubes were opened. When the sulfur dioxide had completely evaporated, they were rinsed out with some carbon tetrachloride which had been purified by freshly distilling in a tin-glass-air system followed by fractional crystallization. The tubes were now charged with some of the purified carbon tetrachloride, which has a surface tension very close to the value for sulfur dioxide a t 20'. The tubes were now suspended in the thermostat and readings were made of the capillary rise. Assuming Richards and Carver's5 value for the surface tension of carbon tetrachloride to be correct, the radii of the capillaries were calculated Richards and Carver, THISJOURNAL, 43, 845 (1921).

412

Vol. 51

W3RNON M. STOWE

from the equation r = 2y/hdg, where r is the radius of a capillary, h is the capillary rise corrected as shown below, d is the density of carbon tetrachloride a t 20°, g is the acceleration due to gravity and y is the surface tension of carbon tetrachloride a t 20”. The values so found for the radii are listed in Table I. A second method was employed to determine the radii. A thread of mercury was introduced into each capillary tube and the length of the thread was measured by a micrometer in conjunction with a steel decimeter scale. Incidentally, a check on the uniformity of bore of the capillaries was made by moving the thread along. The length remained practically constant in all cases. The radii were calculated in the obvious manner, after using a microbalance to weigh the mercury by difference. These values are given in the third column of Table I. The radii were determined finally by breaking the capillaries off squarely and examining them under a microscope fitted with a micrometer. These results are listed in the fourth column of Table I. The mercury method was probably the most exact and these values are used. The fifth and sixth columns give the percentage of deviation from the mercury values. The author did not examine the probable error of these methods. The agreement between the sets was considered very satisfactory. TABLEI RADIIO F THE CAPILLARY TUBBSIN CENTIMETERS Tube

1 2 3 4 5

6

ISt,

CCh

0.01978 .01840 .00923 .01692 .02677 .01409

Method 2d, H g

3d, ocular

0.01950 .OB632 .009135 .017117 .02737 .014222

0.01995 .01868 .00923 .01736 .02729 .01430

-Deviation1% %

$1.4 +1.3 +1.0 -1.2 -0.5 -0.9

3d, %

4-2.3 $0.3 +1.1 $1.4 -0.3 $0.6

Results The calculations were made in the same manner as those performed by Quinn in the article mentioned. Table I1 lists the temperature, the tube number, the observed height, the height after two corrections had been made, the densities of the liquid and of the gas, the difference between these values, the surface tension and the average of these numbers a t each temperature. The corrections were for volume of the meniscus, and a correction for the effect of the outer tube, made in the same manner as Quinn made them. The densities were read from a large graph upon which had been plotted all the data of densities listed by “International Critical Tables.” They were extrapolated mechanically where necessary. Tube No. 3 was thrown out entirely because it was not properly filled with the liquid.

Feb., 1929

THE SURFACE TENSION OF LIQUID SULFUR DIOXIDE

SURFACE 1

-79 -79 -84 -79 -81 -50 -50 -50 -50 -52.3 -20.3 -20 -19.8 -20.2 -20.3 -10 -10 -10 -10 -10 - 5.4 5.9 5.8 - 6.1 5.8 0.2 0.15 0.20 f 0.30 f 0.3 + 5 f 5 + 5 f 5 + 5 9.9 9.9 9.9 9.9 9.9 15 15 15 15 15 20 20 20

-

+ + +

+

TABLEI1 TENSION OF SULFUR DIOXIDE

Tube

hi

ha

1

2.185 2.51 3.05 1.89 3.62 2.465 2.04 2.76 1.705 3.300 2.10 2.22 2.35 1.48 2.90 1.965 2.10 2.27 1.405 2.735 1.93 2.05 2.20 1.36 2.70 1.86 1.975 2.14 1.34 2.57 1.82 1.91 2.08 1.30 2.505 1.76 1.87 2.02 1.26 2.445 1.705 1.795 1.96 1.225 2.375 1.67 1.755 1.90

2.258 2.589 3.138 1.980 3.705 2.546 2.106 2.840 1.787 3.377 2.170 2,291 2.419 1.553 2.969 2.031 2.167 2.337 1.474 2.800 1.995 2.116 2.265 1.428 2.765 1.923 2.039 2.203 1.407 2.632 1.882 1.972 2.142 1.365 2.566 1.820 1.931 2.080 1.323 2.504 1.763 1.853 2.019 1.287 2.433 1.727 1.812 1.956

2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1

2 4 5 6 1 2 4

413

D 1.615

d

D - d

0

1.615

1.5572

0 * 002

1.555

1.4846

,003

1,482

1.4601

.003

1.457

1.4476

.003

1.445

1.4350

.004

1.431

1.4223

.004

1.418

1.4095

,0045

1.4050

1.3964

,005

1.3914

1.3831

.0055

1.3776

7

(34.07) (38.19) 42.53 42.91 41.72 37.85 (29.91) 37.06 37.29 36.60 30.74 31.00 30.08 30.88 30.68 28.29 28.84 28.57 28.82 28.45 27.56 27.93 27.46 27.69 27.86 26.30 26.65 26.45 27.02 26.87 25.51 25.54 25.49 25.97 25.37 24.45 24.78 24.52 24.94 24.53 23.45 23.55 23.58 24.03 23.60 22.75 22.80 22.61

Av. y

30.68

28.69

27.70

26.66

25.58

24.54

23.64

414

Vol. 51

VERNON M. STOWE

TABLEI1 (Concluded) I

Tube

20 20 24.8 24.8 24.9 24.9 24.9 30 30 30 30 30 +40 40 40 40 40 50 50 50 50

5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5 6 1 2 4 5

hi

1.18 2.29 1.59 1.70 1.84 1.14 2.21 1.54 1.63 1.77 1.105 2.15 1.44 1.515 1.65 1.02 1.975 1.32 1.395 1.515 0.94

D

h3

1.240 2.346 1.655 1.756 1.895 1.198 2.264 1.593 1.684 1.823 1.162 2.203 1.490 1.565 1.70 1.073 2.023 1.366 1.442 1.561 0.99

D - d

d

Y

1.3695

.0060

1.3635

1.3556

.0070

1.3486

1.3264

,0100

1.3164

1.2957

,0144

1.2813

22.92 22.54 21.57 21.87 21.68 21.42 21.52 20.54 20.75 20.63 21.03 20.72 18.75 18.82 18.78 18.95 18.57 16.73 16.88 16.78 17.02

Av.

y

22.72

21.61

20.73

18.77

16.85

Discussion The equation of de Block mentioned by Quinn was applied to the data between -20" and +SO" and found to fit within 0.5%. This equation states that y = K(t, - t)", where t is the temperature concerned, in degrees centigrade, t, is the critical temperature, y is the surface tension and K = y J ( t c ) " , Here y ois the surface tension a t 0' and n is a constant, equal to 1.2 for most normal unassociated liquids. In the present work n was determined for each experimental datum and the values were averaged, yielding 1.19. However, the median value seemed more desirable on inspection. This value for n is 1.20. The surface tension was therefore TABLE111 CALCULATED RESULTS t , -2.

-20 - 10

-

5.8 $ 5 9.9 4-15 +20 24.5 30 40 50

+

-

Yoalod, it

Yoba.

1.17 1.13 1.06 1.28 1.27 1.20 1.20 1.22 1.19 1.20 1.20

30.68 28.59 27.70 25.58 24.54 23.64 22.72 21.61 20.73 18.77 16.85

(n

Diff.

1.2)

30.70 28.70 27.84 25.65 24.66 23.64 22.65 21.69 20.69 18.76 16.86

$0.02 .11 .14 .07 .12 0 .07 .08 .04

+ + + + -

+

.01

+ .01

-

Yoalod.

(n

1.19)

30.74 28.69 27.83 25.66 24.68 23.67 22.68 21.73 20.73 18.81 16.92

Diff.

$0.06

+ .lo + .13 + -08 + .14 + .03 - .04 + .12 0 + .04 + .07

Feb., 1929

THE SURFACE TENSION OF LIQUID SULFUR DIOXIDE

415

calculated from both equations, y = 0.061534(157.5-t)1.2 and y = 0.06473 (157.5-t)'-19. The results are entered in Table 111, with the deviations of each from the experimental values. The value n = 1.20 appears to fit slightly better than n = 1.19, although the latter yields smaller maximum deviations. The critical temperature was taken from the work of Cardoso and Fiorentina.6 The functions mentioned by Hildebrand,3in which E, = the total energy of surface formation and v = molal volume, were evaluated as follows: E , = 79 a t 20'; y / ~ "=~6.33 a t 20"; EO/v'/' = 22 a t 20". The second function classifies sulfur dioxide between chloroform and benzene in the Hildebrand solubility table. Numerical values were substituted in the equation of Ramsay and Shields.' 71

(M/d1)'/8 tz

YZ

(M/d,)'/3

- tl

= k

The data for -20 and +50° place the value of k a t 2.134. Non-polar liquids average about 2.12. This work represents a repetition and an extension of crude experiments reported before the North Dakota Academy of Science on May 4, 1928. The author wishes to express his thanks to Dr. S. C. Lind, Director of the School of Chemistry of the University of Minnesota, for his courtesy in placing facilities of the School a t his disposal. Summary

1. Measurements of the surface tension of liquid sulfur dioxide are reported a t temperatures from +50 to -20" with less accurate measurements down to -80" in the undercooled region. 2. The equation of de Block in the form y = 0.061534(157.5-t)1.2 is found to fit the experimental data excellently over a 70" range. 3. Several constants depending upon the value of the surface tension are calculated. Results classify liquid sulfur dioxide with benzene and chloroform in its solubility relations. MINNEAPOLIS, MINNESOTA ~~

Cardoso and Fiorentino, J . chim. phys., 23,841 (1926). 7 Getman, "Outlines of Theoretical Chemistry," John Wiley and Sons, Inc., New York, 1918, p. 148. 6